[RLlib] APPO enhancements (new API stack) vol 03: Fix target network update setting and logic. #48802
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sven1977
added
rllib
simonsays1980
approved these changes
Nov 19, 2024
| and `train_batch_size_per_learner=500`, then the target net is updated | ||
| every 2*4*2*500=8000 trained env steps (every 16 batch updates on each | ||
| learner). | ||
| The authors in [1] suggests that this setting is robust to a range of |
There was a problem hiding this comment.
Awesome. These comments are so helpful!
…_freq Signed-off-by: Sven Mika <svenmika1977@gmail.com>
…net_update_freq' into appo_enhancements_03_fix_target_net_update_freq
…ments_03_fix_target_net_update_freq # Conflicts: # rllib/algorithms/appo/appo.py
…_enhancements_01_circular_buffer
…ments_03_fix_target_net_update_freq
jecsand838
pushed a commit
to jecsand838/ray
that referenced
this pull request
Dec 4, 2024
…update setting and logic. (ray-project#48802) Signed-off-by: Connor Sanders <connor@elastiflow.com>
dentiny
pushed a commit
to dentiny/ray
that referenced
this pull request
Dec 7, 2024
…update setting and logic. (ray-project#48802) Signed-off-by: hjiang <dentinyhao@gmail.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
APPO enhancements (new API stack) vol 03: Fix target network update setting and logic.
NUM_ENV_STEPS_TRAINED_LIFETIMEmetrics, not SAMPLED(!)nfrom the paper, where n=N [number of batches in buffer] * K [max iterations per batch] * [train batch size].target_network_update_freqfrom 1 to 2. The current setting of 1 makes it such that the network gets updated after each batch update, b/c the unit used (individual timesteps) is wrong.Why are these changes needed?
Related issue number
Checks
git commit -s) in this PR.scripts/format.shto lint the changes in this PR.method in Tune, I've added it in
doc/source/tune/api/under thecorresponding
.rstfile.